Solution diverges/Unstable
 

Problem: Turbulent flow in a duct with a hole on the surface.

I created the trimmed mesh in proam. While writing the geom file it gives me the following warnings.

-----------

**WARNING - NO CELL FOUND FOR BOUNDARY 4531 IN REGION 1 BOUNDARY SET CLEARED **WARNING - NO CELL FOUND FOR BOUNDARY 4548 IN REGION 1

**WARNING - 2 BOUNDARIES NOT FOUND ON ANY CELL IN THE MODEL.

THE BOUNDARY SET NOW CONTAINS ONLY THESE BOUNDARIES.

----------

The solution diverges when I try to do the analysis...

How do read the new file.div? Does it have any information?

Thanks in advance

Radhika

Re: Solution diverges/Unstable
 

The new.div file is the same as a .pst file and can be post processed in the same way. To be honest I have never gotten any help from it.

I doubt if your error messages have anything to do with the problem. Next time you write the geometry file, just delete the bset written.

When does the solution diverge; i.e., how many iterations. If it diverges quickly, then it is probably a mesh problem. Do you have any warnings in your .info file about centriods? I have found that a major problem that I run across is crack errors.

If you are getting a ways (100 iterations or more) try under-relaxing variables. Remember, CD is more stable than AMG and UW is more stable than MARS.

Tom

Re: Solution diverges/Unstable
 

Hi

The solution is diverges after 10 iterations. But when I run the check for the mesh I do not see any errors, like cracks etc...just the warning I mentioned previously..

Also, I have the following warnings in the .info file.

----

WARNING #042 *** PROCESS NOT FULLY CONVERGED -- EQ.,NIT,RESI: DISS 1nan

*** WARNING #045 *** SOLUTION DIVERGES; RESCUE NOT ATTEMPTED; RETURNING OLD SOLUTION

----

What is DISS 1nan?

I used bset, , , , , , 0 to delete the bset. Is that right?

Regards

Radhika

Re: Solution diverges/Unstable
 

To delete the boundaries, run the boundary check. Afetr you get the message that says they were put in the boundary set do

bdel, bset

nan means not a number.

One thing to check is that the inlet sees the outlet. Use the check tool gui to check for cracks and connectivity.

Use CD as your solution method and and UD because these are the most stable. Also, if you can, start the analysis using manual intialization.

If none of these work, try decreasing your flow rate some to see if the resistance of your system is too high for your flow.

Good luck.

Tom

Re: Solution diverges/Unstable
 

Thanks a lot!

The solution is converging! I just had to get rid of the warnings..